A Human–Machine Interaction Mechanism: Additive Manufacturing for Industry 5.0—Design and Management

Abstract

Industry 5.0 is an emerging value-driven manufacturing model in which human–machine interface-oriented intelligent manufacturing is one of the core concepts. Based on the theoretical human–cyber–physical system (HCPS), a reference framework for human–machine collaborative additive manufacturing for Industry 5.0 is proposed. This framework establishes a three-level product–economy–ecology model and explains the basic concept of human–machine collaborative additive manufacturing by considering the intrinsic characteristics and functional evolution of additive manufacturing technology. Key enabling technologies for product development process design are discussed, including the Internet of Things (IoT), artificial intelligence (AI), digital twin (DT) technology, extended reality, and intelligent materials. Additionally, the typical applications of human–machine collaborative additive manufacturing in the product, economic, and ecological layers are discussed, including personalized product design, interactive manufacturing, human–machine interaction (HMI) technology for the process chain, collaborative design, distributed manufacturing, and energy conservation and emission reductions. By developing the theory of the HCPS, for the first time its core concepts, key technologies, and typical scenarios are systematically elaborated to promote the transformation of additive manufacturing towards the Industry 5.0 paradigm of human–machine collaboration and to better meet the personalized needs of users.

1. Introduction

Industry 5.0 is an initiative to promote “artificial intelligence” at the technical level and “social 5.0” at the social development level, representing a reflection on the current intelligent manufacturing paradigm or Industry 4.0. Since Industry 1.0, industrial development has been committed to improving people’s lives, whereby various production organizations engage in user-centered value creation, solving problems for users or making problem-solving processes more efficient. The sudden COVID-19 epidemic and the changes in international patterns have highlighted the importance of the elasticity and robustness of the enterprise supply chain. The worsening climate and ecological environment have exacerbated people’s concerns about fossil fuels, environmental pollution, ecological imbalances, and other issues. Since Industry 4.0, the emergence of unmanned production lines has sparked people’s reflection on the prospects of “less human” and “unmanned” industries. Industry 5.0 proposes solutions to the above issues, placing greater emphasis on social and ecological values and shaping the direction of industrial development from a value system to a value-driven industrial model based on Industry 4.0 [1,2,3]. The white paper on Industry 5.0 released by the European Union states that Industry 5.0 has three characteristics, namely being “human-centric,” “sustainable,” and “resilient,” reflecting Industry 5.0’s responses to the issues mentioned earlier [4]. Value-driven Industry 5.0 consists of a value network of five elements—efficiency, safety, humans, machines, and nature (see Figure 1). The benefit is reflected in the personalized value of products and services. Its efficiency is reflected in the effectiveness of the low inputs and high outputs in the production and operation processes. Its safety is reflected in the resilience of the enterprise supply chain. The human–machine approach is reflected in the people-centered value of human–machine cooperation; naturally, this is reflected in the sustainable value of enterprises to the Earth’s environment. The system’s value includes all of these aspects.

In fact, worldwide manufacturing industries from Industry 1.0 to 5.0, even within the same enterprise, can exist in a state of mass production, mass custom production, and mass personalized production (“longtail manufacturing”) at the same time (as shown in Figure 2). No matter which manufacturing model or models, the demand is centered around people. People are both the manufacturing participants and demand presenters [5,6,7]. To this end, in response to the Industry 5.0 trend proposed by the EU, China’s leading manufacturing enterprises can start reactive strategic preparation initiatives covering at least three aspects—their competitive strategy, multi-space synergy strategy, and internationalization strategy.

Industry 5.0 aims to put human well-being at the center of the industrial manufacturing system, strives to achieve social goals beyond employment and growth, and provides strong support for the sustainable development and prosperity of all people, sounding the clarion call for a new round of global industrial transformation [8,9]. Unlike Industry 4.0, which is primarily technology-driven, Industry 5.0 is more value-driven and includes three core concepts—a people-centric approach, sustainability, and resilience. The objective of the study conducted by Narkhede et al. [10] was to comprehensively examine the correlation between I5.0 and sustainable manufacturing, identify its effects and implementation difficulties, analyze its influence on triple bottom line sustainability, and propose a comprehensive framework for the implementation of I5.0 in Indian manufacturing enterprises.

With human-centered manufacturing, the ultimate goal of manufacturing is to promote human well-being, advocating that the different participants in the manufacturing value chain should be valued. Shoukat et al. [11] clarified the key concepts of human-centered intelligent manufacturing (human-oriented intelligent manufacturing) based on the theoretical HCPS and established a research framework and technical system. Mudassar et al. [12] suggested a smart manufacturing system that utilizes DT technology to enable traditional project-based companies to efficiently gather, store, and process data. This system incorporates AI algorithms to facilitate decision-making in project selection, scheduling, and resource allocation processes. Mumtaz et al. [13] examined the current state of the printed circuit board manufacturing environment and explored future advancements in the direction of smart manufacturing systems. They introduced a cloud manufacturing paradigm in the current smart manufacturing environment to provide convenient access to resources, materials, manufacturing processes, planning issues, and data sharing networks.

Recently, research on human-centered manufacturing involving specific manufacturing processes, technologies, and machines has been gradually emerging, although it is still in its infancy. In the evolution from Industry 4.0 to Industry 5.0, in depth analyses of the theoretical system, enabling technologies, and specific scenarios of human centricity at the level of specific manufacturing processes will greatly improve the understanding of human–machine oriented manufacturing for both academia and industry and promote its application in subdivided fields.

As an important development direction for advanced manufacturing and a core component of intelligent manufacturing systems, additive manufacturing technology is a transformative technology with a wide range of application prospects, which can be used to efficiently and reliably implement complex design schemes [14]. In the early stage of development, additive manufacturing technology was mainly used as a rapid prototyping tool for product development and concept design, assisting designers to materialize and visualize design solutions. At the same time, the development of additive manufacturing machines, both software and hardware, is profoundly reshaping the relationship between humans and machines [15]. With the development of technology and changes in the human–machine relationship, HMI additive manufacturing for Industry 5.0 is facing new challenges:

• Due to the growing importance of additive manufacturing in product development and the range of application scenarios, more theoretical research is needed to understand the complex HMI.
• HMI-oriented manufacturing must take into account the unique requirements of each stage of product development.
• The unique features of additive manufacturing, such as digitization and layer-by-layer addition, avoid the direct application of HMI laws and methods used in other manufacturing technologies.

Instead, the further development of enabling technology for Industry 5.0 is necessary, based on a comprehensive understanding of the characteristics mentioned above. In order to achieve more user-friendly and efficient human–robot collaboration opportunities, it is necessary to analyze the interactions between humans and additive manufacturing machines and understand the specific interaction needs of different links. Hence, it is crucial to investigate the collaboration opportunities between humans and machines in order to fully harness the capabilities of additive manufacturing and achieve the fundamental principles of Industry 5.0—production that prioritizes human needs.

• This study involved materializing the concept in the conceptual design stage, co-creating and customizing in the detailed design stage, making collaborative decisions in the manufacturing stage, and repairing and remanufacturing parts in the maintenance stage.
• This paper presents a theoretical structure for cooperative additive manufacturing in the context of Industry 5.0.
• It explains how HMI affects the process of product development and the essential technologies involved.
• It also outlines how this study can be applied in typical situations and discusses the challenges associated with it.

2. Preliminary Information

2.1. Concept of HMI Additive Manufacturing

According to the theoretical framework of the HCPS [4], HMI additive manufacturing for Industry 5.0 involves a human-centered approach to additive manufacturing. This approach aims to better fulfil user requirements and contribute to sustainable social development. In this study, a three-level model of HMI additive manufacturing for Industry 5.0 is constructed from the perspectives of the product layer, economic layer, and ecological layer, as shown in Figure 3.

The model takes the HMI relationship in additive manufacturing as the research object and is composed of three elements—humans, machines, and the human–machine collaborative relationship. Humans include designers, engineers, and users, among others; machines include additive manufacturing machines and digital design software; and the HMI connection is the relationship between humans and machines in all aspects of product development. At the product layer, the HMI relationship is marked by diversification, parallelism, and dynamics. The economic layer introduces new models such as multi-time and space-collaborative innovation models and cross-regional collaborative manufacturing through the use of additive manufacturing services. At the ecological layer, the human–machine collaboration relationship exhibits characteristics of multi-disciplinary and multi-field intersection with the advancement of the “additive manufacturing + complex sociotechnical system” paradigm.

As the function of additive manufacturing in the product development process changes and the application scenarios diversify, the HMI relationship at all levels becomes more complex. Additive manufacturing, unlike subtractive and isobaric materials, is a digital-technology-driven method that allows for precise control over the fineness and resolution of raw materials. This enables the production of three-dimensional entities with high-density shape information and high-resolution physical property information. As a result, the shapes and properties of products can be defined simultaneously [16]. Compared with other manufacturing scenarios such as HMI assembly and HMI disassembly, the role division process in HMI additive manufacturing is different and the characteristics of the HMI relationship are more prominent, emphasizing that humans and machines participate in the manufacturing process together, cooperate with each other, and jointly create products. In addition, HMI additive manufacturing operates in a different way, highlighting the collaborative characteristics in the virtual space and emphasizing the real-time feedback mechanism. Humans can obtain real-time feedback through interactions with machines in virtual and real spaces to make adjustments and optimize processes. In HMI assembly and disassembly, people use more hands or tools to interact directly with physical objects. Therefore, it is not enough to simply migrate the HMI relationship into an automated shop floor for other manufacturing technologies to achieve additive manufacturing.

With the continuous advancement of HMI technology, the boundaries of the HMI relationship in additive manufacturing in the virtual and real spaces are more blurred than in the past and the methods of human–computer co-creation are more abundant, making it necessary to re-examine the traditional collaboration method between humans and additive manufacturing machines. The research on HMI additive manufacturing involves many scientific issues, such as how to achieve HMI planning and control to improve the efficiency and quality of the additive manufacturing process [17] and how to design intelligent interfaces and interaction methods so that humans and machines can communicate and collaborate efficiently through multi-modal interaction [18]. These scientific questions need to be studied in depth to advance the development of HMI additive manufacturing. In the product layer, economic layer, and ecological layer, the human–physical system, human–information system, and physical–information system undertake a variety of tasks for different purposes. Therefore, it is of great significance to sort out and analyze the characteristics of the HMI relationship at each level to understand and continue to explore human–machine collaborative additive manufacturing for Industry 5.0.

2.2. Characteristics of HMI Additive Manufacturing

2.2.1. Product Layers

From the perspective of the product layers, additive manufacturing technology has the characteristics of being flexible and agile and allowing digital manufacturing, and its HMI relationship is closely related to these characteristics. The traditional HMI additive manufacturing process involves one-way interaction and the separation of virtual and real spaces. For example, additive manufacturing was first used in the concept design stage in product development to quickly complete concept materialization and verification testing processes. Designers use computer-aided design software to plan the subsequent processes and complete the physical manufacturing work. In this process, the interactions between humans and the additive manufacturing system are virtual and real and the mode of the HMI relationship is serial. With the support of digitization, networking, and intelligent technologies, the human–machine relationship in additive manufacturing at the product level has taken on new characteristics.

(1)

Diversification, also known as HMI in additive manufacturing, is a diverse model that combines active, passive, and dual active modes. For designers, the development of machine intelligence has enabled machines to become active tools for intelligent design, which can be used in exploring high-dimensional design spaces, searching for high-performance design solutions, and achieving multi-objective optimization through the computing power of machines [19]. At the same time, machine intelligence also empowers batch additive manufacturing by using network ontology language to semantically model and retrieve design knowledge for additive manufacturing, assisting in the process planning for additive manufacturing [20].

(2)

Parallelization is where the HMI relationship is integrated into the virtual and real spaces and human–machine collaboration occurs simultaneously throughout product creation. Due to its digital nature, additive manufacturing with HMI relies heavily on virtual spaces. Virtual and actual space integration improves the HMI and product development processes. Extended reality technology allows designers to perceive virtual and physical spaces from multiple dimensions. Virtual haptic feedback can be used to improve perception, provide timely feedback, and help designers adjust the virtual space design parameters to meet users’ personalized needs [21].

(3)

Dynamic additive manufacturing has evolved from single-machine and single-process manufacturing concepts to a reliable and efficient production system involving multi-machine collaboration and multi-process manufacturing, while the HMI relationship is in a state of high uncertainty and dynamic change. The human–machine collaboration relationship faces new challenges when manufacturing and post-processing complex products with different quantities and styles, such as for the real-time acquisition, processing, and storage of heterogeneous information, including product, user, and machine status data.

2.2.2. Economic Layer

From an economic perspective, additive manufacturing services are creating new production and consumption patterns, covering the design, production, and consumption processes, such as creative design, distributed production, additive manufacturing, and customized consumption, as shown in Figure 4. Single-machine and multi-person, multi-machine collaboration concepts involve spatiotemporal separation and system intersection. Crowdsourcing is an organizational strategy that uses the Internet to distribute professional designers’ product design work to the public for free [22]. In this approach, data flows connect humans, additive manufacturing information systems, and physical systems. Additionally, this pattern is still difficult to apply to complicated product design processes, such as designing human–human collaboration platforms and human–machine collaborative strategies for multi-temporal and multi-spatial group creative design. Increased design innovation can also change production models. The process of distributed production with additive manufacturing involves globally dispersed additive manufacturing facility networks and big data analysis capabilities [23].

The evolution of this model requires the following core issues to be addressed, including the development of seamless digital production workflows and agile order management systems. Distributed manufacturing is beneficial in improving the resource waste outputs in the design and production process in traditional manufacturing industries, such as the high opening costs and transportation and logistics resources required for raw materials and products. At the same time, this model also helps to promote the development of the customized consumption concept and meet the personalized needs of users at a lower cost. Customized consumption refers to people purchasing products tailored to their own needs [24]. The development of this consumption model relies on the low-cost and reliable manufacturing capabilities provided by distributed manufacturing systems based on additive manufacturing, as well as the product development process for HMI additive manufacturing. The increasing demand for personalized products from consumers further drives the trend of collaborative design towards achieving automation, high product quality, and strong innovation results. The maturity and development of the above new models provide opportunities to meet users’ needs and realize the value of humans, and the deep integration of the three models will further promote the transformation of production and consumption patterns.

2.2.3. Ecological Layer

From the perspective of social ecology, the integration of HMI additive manufacturing and complex social technological systems is creating new value for the sustainable development of a harmonious society and the harmonious coexistence between humans and nature, as shown in Figure 5.

In terms of social livelihood, the “additive manufacturing + education” concept provides a teaching model for students in economically disadvantaged areas at a lower cost [25], which helps to achieve educational equity for minority groups. The “additive manufacturing + healthcare” concept involves the provision of customized medical aids and artificial implants for patients at lower costs and faster speeds, which helps to achieve personalized healthcare goals [26,27]. In terms of natural ecology, the “additive manufacturing + carbon neutrality” concept reduces the wastage of raw materials in the production and manufacturing processes and provides the conditions for producing more efficient and clean energy [28], which helps to achieve the goal of carbon neutrality. The “additive manufacturing + smart manufacturing” concept is a combination of manufacturing processes that offers digital sustainability.

2.3. HMI Additive Manufacturing-Based Product Development

The traditional product development process generally consists of successive phases, which include concept generation, design, prototyping, testing, and manufacturing. Although this approach has played a fundamental role in production, its linear structure imposes inherent limitations, sometimes resulting in time-consuming iterations and inflexible design constraints. On the other hand, HMI additive manufacturing provides a flexible and efficient option that combines human ingenuity and machine abilities to transform the process of developing products. HMI additive manufacturing enables designers to directly engage with the manufacturing process, promoting iterative refinements and rapid innovation through the utilization of modern technologies such as 3D printing and augmented reality. In order to meet the personalized needs of different stakeholders, this study proposes a product development process based on HMI additive manufacturing, as shown in Figure 6.

It can be seen in Figure 6 that the process considers the concepts of maintenance manufacturing, conceptual design, and detail design with the DT-based merging of information and physical systems. This model allows data transfer between the user, design staff, and engineer. It aims to integrate human and machine capabilities to achieve improved HMI and creativity. The method considers the varied requirements of multiple groups of individuals, including designers, developers, and users, and tailors the interface accordingly, with a focus on human needs. Simultaneously, the selection process of the active and passive relationships between humans and machines takes into account the interaction features of individuals and virtual and physical places in various contexts. Typically, the interaction between humans and information systems in the conventional development process involves the use of a computer-aided design tool that offers enhanced modelling, analytical, and decision-making capabilities. The approach facilitates a dynamic interface that enables machines to utilize AI to stimulate humans to participate in cognitive tasks, including creative and integrated decision-making tasks. Furthermore, humans have the ability to intervene in the process of developing additive manufacturing products either directly or indirectly through the use of sensors or smart materials. Additionally, they can utilize DT technology to facilitate interactions between information systems and physical systems [29].

The traditional product development process is a serial model, in which the designer enters the initial design concept and completes the concept materialization and verification testing in the concept stage with the help of additive manufacturing resources, then the designer and digital design software complete the detailed design process and the digital model and process planning parameters are input into the physical additive manufacturing system to complete the production and manufacturing of the product. The human-centric values of Industry 5.0 are evolving the modern product development model towards a parallel model. In the process of product development, designers can instantly and fully obtain data and knowledge from the product, user, and production sides and use data-driven methods for the personalized design of products. Therefore, the product development process based on human–robot collaborative additive manufacturing needs to take into account the needs of people in the design, manufacturing, and use of the product, and the key to success lies in the use of data. In addition, HMI additive manufacturing allows for the smooth incorporation of DT technologies, which aid in the virtual prototyping and predictive analysis processes to proactively detect and address design flaws. This predictive capability not only decreases the time it takes to bring a product to market but also reduces the need for expensive revisions, optimizing the use of resources and improving the overall quality of the product.

3. Key Technologies for Additive Manufacturing

In the three-level reference model of HMI additive manufacturing, the ecological layer and the economic layer cover a wide range of technologies, including a variety of disciplines outside the engineering field, so this study will not discuss them in detail. This section focuses on different engagement models at the product level, proposing key technologies for human-oriented, machine-oriented, and HMI to enable smooth human–machine collaboration in design and manufacturing, as shown in Figure 7.

3.1. Internet of Things

The IoT facilitates the process of connecting objects and devices in physical spaces with data in order to achieve interconnections between people, objects, data, and processes. The data generated by the IoT help users to obtain customized products and experiences, which can lead to improved user satisfaction. The scope of the IoT extends to interactions between human and machines, humans and humans, and humans and the environment, which helps in harnessing the intelligence of additive manufacturing products [30,31]. The IoT can remove information barriers from the production process, reduce operating costs, and lead to improved production efficiency. Additive manufacturing systems uses standard communication protocols for data exchange to ensure the efficiency and security of the HMI systems. For example, Liu et al. [32] used the MTConnect protocol to obtain device information and adopted the OPCUA protocol for cloud edge communication. The data generated by the IoT is transmitted through 5G communication technology with ultra-low latency and high reliability rates, providing value in the human-oriented product development process. During the product design phase, user preferences can be incorporated into the conceptual design of the product by analyzing Internet data. During the manufacturing stage of a product, the IoT can be used to monitor the manufacturing process in real time. It can also be used to intelligently analyze the data created during the process to detect and maintain equipment as needed. During the consumption phase of a product, a range of sensors cab be used to to gather data during the usage process, creating a data loop that enables the product to become very intelligent and facilitates iterative development.

3.2. Artificial Intelligence

AI technology, as a powerful tool for automated decision-making and for solving high-dimensional complex problems, has made significant contributions to many fields of Industry 5.0 and plays an important role in the entire product lifecycle. However, in the field of additive manufacturing, the application of artificial intelligence is still in an emerging stage. The main challenge faced by AI systems in additive manufacturing is how to utilize computing and perception capabilities that are greater than those of humans to assist human decision-making and meet the additive manufacturing needs in different scenarios. This requires the AI systems to consider resources in different scenarios, such as the data type, data quality, and algorithm’s robustness. On the design side, AI can be used to assist designers in the design analysis and synthesis processes, such as for the lightweight design of vehicle bodies [33] and generation of customized product ideas. AI can be used to automatically create design variants by processing heterogeneous and unstructured data, including computer-aided design models, text, images, and video information. Designers shift their focus from the entire design process of generating creative ideas to becoming more focused on design tasks such as planning, decision-making, and communication. The challenges brought about by this transformation include how designers can pose better questions to artificial intelligence systems, how AI can be used for improved design efficiency and quality, and how AI can be applied for enhanced communication efficiency among design teams [34]. On the manufacturing side, AI can be used to assist engineers in selecting and planning manufacturing processes [35].

Engineers can directly engage in natural language conversations with machines through the use of large language models such as ChatGPT, reducing barriers and misunderstandings during information transmission. This makes the process planning and parameter adjustment processes in additive manufacturing more efficient and accurate, improving the efficiency and quality of the manufacturing process. However, AI systems face challenges relating to data scarcity and access barriers. In most commercial additive manufacturing machines, the data environment is usually closed, which means that data acquisition is difficult and the dataset available for AI training is small. To address these issues, cutting-edge research studies have applied small sample learning [36] and adversarial neural networks. In the use and maintenance of products, AI helps additive manufacturing machines achieve intelligent detection, predictive maintenance, and multi-machine collaboration in the production process. This process requires AI systems to make predictions and decisions quickly and accurately to avoid machine failures. Cloud computing [37] and edge computing provide opportunities to solve these problems. Threats to the security of AI systems can come in many forms, such as hostile attacks, data breaches, and intentional manipulation. AI algorithms and models can be exploited to deceive or control automated systems, which can have severe implications. Strong cybersecurity mechanisms, such as encryption, authentication, and anomaly detection, are essential for protecting AI systems from security threats. The swift progress of AI technology has surpassed the existing regulatory frameworks and legal requirements, presenting governance and compliance difficulties. However, AI systems have low interpretability capabilities, which poses challenges in additive manufacturing applications with high security requirements for HMI systems.

3.3. Digital Twin

DT technology integrates multiple physical and multi-scale attributes with real-time interaction and high-fidelity characteristics to map and fuse digital spaces with physical spaces [38]. With the development of the IoT and the generation of communication technology, data from the physical space can be fed into the digital space for simulation. This method of mapping real-time objects digitally through DT systems makes it possible to analyze and monitor digital objects and prevent physical object problems. DT technology has played a crucial role in the production and manufacturing of products in intelligent workshops and automated production lines, as well as in inspection and maintenance processes. At the same time, DT technology and the digitization characteristics of the additive manufacturing process complement each other and play an important role in innovative product design.

Designers can continuously improve the digital design models in virtual spaces by utilizing feedback from user interactions, which can then be fed back into the physical space of the product [39]. From the perspective of HMI additive manufacturing, the development of DT technology presents the following challenges. Firstly, multi-physics field simulations present a challenge. In additive manufacturing, models with different physical properties need to be associated together to accurately simulate, diagnose, predict, and control the digital space, which is the key to achieving intelligent detection in additive manufacturing. Secondly, complete process simulations are also challenging. Digitally mapping the real-time data collected by various sensors in each stage of product development helps designers better understand the dynamic needs of users and the dynamic changes in product performance throughout the entire lifecycle. Thirdly, achieving an unobstructed data flow is challenging. By combining the inherent digital characteristics of the additive manufacturing approach and bridging the breakpoints in the data flow during current product development, this will be conducive to achieving a continuous and bidirectional data flow in the product development process, thereby achieving more friendly and efficient human–machine collaboration and creation outcomes.

3.4. Extended Reality

Extended reality is a hybrid of virtual reality (VR), augmented reality (AR), and mixed reality (MR) technologies. In Industry 5.0, AR technology plays an important role in many different fields, such as remote assistance [40], telemedicine [41], and so on. For additive manufacturing, AR technology can be used to bridge the virtual and physical worlds together, further amplifying the digital characteristics of the additive manufacturing approach, reducing product development times, and meeting customization needs. Brain–computer interface (BCI) technology has the ability to enhance a human’s cognitive capacities by combining machine intelligence with human creativity and intuition. BCI systems utilize real-time feedback and adaptive algorithms to aid users in brainstorming, ideation, and problem-solving tasks, resulting in the generation of unique solutions and fresh perspectives. At the same time, such technologies can also increase the flexibility of time–space interactions between human and additive manufacturing machines; for example, this could mean that people do not need to complete collaborative work in limited places and during working hours.

Extended reality devices can provide in-depth insights into professional issues for HMI technology during product development, which is conducive to unleashing human creativity and imagination. With the help of AR technology, users can gain a more intuitive understanding of virtual information in real scenes and interact with it, greatly improving the efficiency of engineers in performing complex assembly tasks [42]. Leutert et al. [43] utilized projection devices to achieve remote bidirectional communication with spatial references, allowing for efficient remote maintenance. However, projection devices are limited by their specific viewing angles and environmental requirements, and darker or uneven lighting environments may affect the projection effect and visibility. Extended reality technology has created the conditions for the virtual representation of complex environments in industry, allowing users to interact with machines, products, and the environment faster and more easily, which in turn promotes value creation. In order to achieve better HMI additive manufacturing outcomes, many improvements must still be made to AR technology. For example, the question remains of how to enhance human experiences in both virtual and physical spaces and how to achieve the seamless stacking and fusion of virtual and real spaces. In order to improve the dimensional accuracy of virtual–real fusion initiatives, in addition to using high-precision hardware devices such as sensors, AR systems based on specific processes can also be used. Alternatively, visual tracking technology can be used to enhance the environmental perception ability of extended systems in specific scenes [44]. For different manufacturing tasks, the simultaneous use of multiple devices provides flexibility, overcomes the limitations of the manufacturing environment, and leads to improved user comfort. However, it introduces challenges such as increased cognitive loads for users and complex task planning processes.

3.5. Intelligent Materials

Intelligent materials are digital materials that have the ability to perceive and provide feedback on external stimuli. The perception and feedback functions of smart materials can be used to obtain data on production and usage processes without the use of external sensors. This broadens the HMI interface during use of the product, allows the bidirectional flow of information in the product development process, and improves the iterative development speed of the product. The unique advantage of additive manufacturing lies in having sufficient time and space windows to actively control the fine-grained resolution of raw materials. The output three-dimensional entity not only carries high-density geometric information but also high-resolution material property information, allowing the final product to be formed and qualitatively completed synchronously. Because the process planning strategy for additive manufacturing allows for the construction of multi-dimensional and multi-material architectures, various scales of intelligent materials, from macro- to micro-structures, can be directly controlled through the additive manufacturing process. In order to achieve more dimensions of human–machine collaboration, further development of application scenarios for intelligent materials should be carried out. Firstly, the integrated design process of the material structure, such as the design of functionally graded materials, is aimed at further improving the performance of the product. Secondly, the multi-dimensional feedback from smart materials can be utilized, providing users with interactive feedback such as visual and tactile information. Thirdly, digital design software could be developed for intelligent materials to support designers and engineers in synchronously controlling the geometric shape and performance of raw materials in the virtual space.

3.6. Other Technologies

In addition to the key enabling technologies at the product level, the enabling technologies at the economic and ecological levels are also driving the development of HMI technology in the direction of Industry 5.0. In the information age, the use of big data has become a key factor in customized consumption initiatives, and enterprises use big data to analyze the needs of each user and make strategic decisions through the Internet [45]. With the rapid growth of users’ individual needs, the market’s demand for innovation continues to increase. The combination of IoT and distributed 3D printing technologies lays the foundation for personalized manufacturing, accelerating the shift of the production paradigm to mass customization [46]. New localized and digital supply chains can improve the traceability and responsiveness of supply chains through the use of blockchain technology. Innovations in complex sociotechnological systems have positive impacts on complex social issues such as education, healthcare, transportation, government policies, and environmental protection [47], such as the transformation of engineering education concepts and models, the application of green manufacturing technologies, and the development of recyclable materials. Material recycling involves closed-loop manufacturing techniques that aim to conserve resources by recycling and reusing materials, hence minimizing the environmental effect of the production process. Remanufacturing, refurbishing, and product upgrading processes increase the longevity of items, thereby decreasing the necessity for new production efforts and limiting waste creation. The creation of sustainable supply chains involves collaborating with suppliers to implement environmentally friendly practices, procuring renewable materials, and minimizing transportation emissions. These efforts all lead to the development of a more sustainable manufacturing ecosystem.

4. Typical HMI Additive Manufacturing Scenarios

Based on the three-level model, product development framework, and key technologies for HMI additive manufacturing, this section takes product-level cases involving personalized product design, interactive design and manufacturing for human–machine co-creation, and the use of HMI technology for the additive manufacturing process chain as examples and expounds on the application of key HMI additive manufacturing technologies and the opportunities and challenges faced in the use of HMI technology. At the same time, this section also takes economic cases involving creative crowdsourced design and distributed manufacturing and ecological cases involving energy conservation and emission reductions as examples and analyzes the impacts of key HMI additive manufacturing technologies on these scenarios according to the different needs of people in different scenarios.

4.1. Personalized Product Design

The characteristics of digitization and low manufacturing constraints in additive manufacturing allow for greater design freedom in the product development process, which helps in providing personalized products, services, and experiences for users. For consumer goods that address obvious personalized needs, the collaborative creation relationship between humans and machines has shown significant importance. AI technology can be used to process unstructured data and transform it into design knowledge, enhancing the imagination of designers and increasing user engagement in the design process. For example, natural language processing techniques can be used to analyze user preferences from user comments [48] and adversarial generative networks can be used to design and synthesize fashion images. More cutting-edge research studies could be performed involving multi-modal design synthesis, such as using generative deep learning models to generate images based on text information. The IoT enables the integration of user interaction data with products and user preferences into the early stages of product design process, meaning the personalized experiences of users are fully considered in product design [49]. For example, designers can use natural language processing techniques to analyze information related to product preferences in user reviews and then synthesize personalized seat shapes through the use of adversarial generative networks. Through the collection of contact pressure distribution data through sensor arrays dispersed across the surface of a seat and the subsequent matching of lattice cells with customized structural stiffness parameters based on pressure values, the seat’s customized internal lattice structure can be generated for the user, as depicted in Figure 8 [50]. DT technology gives designers the option to simulate user behavior variances in the virtual space. This gives designers the ability to fine-tune digital models of items in the virtual space so that they may cater to the specific requirements of each consumer.

The personalized product design approach needs to be human-oriented, focusing on human needs and feelings. The IoT and AI technology have promoted the rapid development of personalized product design initiatives, although this has also brought about new opportunities and challenges related to the use of HMI in such scenarios.

(1)

The use of data is the key to successfully achieving product personalization. Obtaining high-quality data is a prerequisite for applying AI to assist in the design process. The main challenges faced during data acquisition are small data volumes, low data quality, and a lack of acquisition interfaces. Ensuring data security can also be a potential challenge. With the transformation of personalized design paradigms, ensuring the security of user or company data will be more challenging.

(2)

Multi-modal learning provides opportunities for understanding and evaluating personalized design processes. Multi-modal data can be used to assist people in better understanding problems and also to facilitate machines to gain a more comprehensive understanding of design problems. Therefore, the key to improving efficiency and quality in personalized product design lies in using AI to enable machines to process multi-modal data and assist designers in making decisions.

4.2. Interactive Manufacturing

Interactive manufacturing is an emerging concept for early product development, with the goal of achieving the synchronous exploration and materialization of design ideas. Unlike the traditional design before construction method, interactive manufacturing involves a parallel workflow that allows users to directly interact with physical workpieces and receive immediate feedback. In interactive additive manufacturing, the combination of digital design and manufacturing workflows with traditional manufacturing processes provides more possibilities for the realization of creative ideas. Additive manufacturing technology has become the main tool for converting virtual designs into physical objects [51]. However, the low manufacturing speeds, limited workflows, and limited interaction modes limit the development of HMI technology for additive manufacturing. To overcome these issues, researchers have proposed low-fidelity molding methods, such as those involving wireframe models and instant printing [52], to accelerate the physical molding process. In addition, by simplifying the workflow and designing a hybrid process for additive manufacturing machines and 3D pens, researchers have created features with different fidelity levels to improve the product development efficiency.

In order to expand the interaction mode, designers have used MR technology to assemble irregular objects using additively manufactured connectors [53]. The virtual rendering of objects is possible using technologies such as AR and MR to provide an interactive experience with immediate feedback [54]. For example, Yamaoka et al. [55] used aerial display technology to overlay images on physical objects, providing real-time previews of printed objects. The use of interactive and adaptive design and manufacturing workflows will allow users to explore more comprehensive design spaces. Mitterberger et al. [56] established a customized augmented reality system for on-site construction, which had the same level of complexity and accuracy as robot manufacturing systems through the use of enhanced manual processes, providing a reference for interactive additive manufacturing systems involving multiple people. Ostrander et al. [57] demonstrated the effectiveness of interactive additive manufacturing using virtual reality technology as a tool for novice designers in education by evaluating the knowledge mastery and self-efficacy of students in additive manufacturing. With the development of related enabling technologies, human–machine collaboration initiatives in interactive manufacturing present new opportunities and challenges.

(1)

The real-time perception and control of the state of physical workpieces are key factors that directly affect the quality of the manufacturing results. The utilization of feedback information such as sensor technology and augmented reality technology data to assess and control the printing process in real time is crucial for ensuring the quality and performance of the final product.

(2)

The selection and optimized design of materials require deep collaboration between humans and AI. Interactive manufacturing initiatives must consider the design constraints, and AI can be used to quickly select materials with the required mechanical properties. It can also be used to assist designers in quickly understanding the properties of materials and their interactions during the printing process. In addition, the subjective evaluation of products is closely related to humans’ perceptions of materials and craftsmanship. Therefore, fully leveraging the advantages of both humans and machines can help achieve efficient material selection and design optimization outcomes.

4.3. Process Chain

At the production stage, additive manufacturing has evolved from standalone manufacturing and single process manufacturing systems to a multi-machine collaborative and multi-process manufacturing system. Multi-machine collaboration involves numerous printers working together, while multi-process approach encompass diversity design and planning and additive manufacturing post-processing efforts. For additive manufacturing with multi-machine collaboration, the main task is for humans and machines to safely and efficiently collaborate in a shared space. Zhang et al. [58] developed the Aerial AM system, which coordinates the construction and scanning of drones in two loops through a distributed multi-agent method, completing the 3D printing and machine task allocation processes in the system. Mitterberger et al. [59] created an HMI system consisting of two robots and two people, which was used to assemble complex wooden structures in a shared digital physical workspace, providing a reference for the collaborative manufacturing of complex structures by multiple people and with multi-degree of freedom printers.

Due to the need for senior engineers to manually correct parts with complex shapes, it is necessary to develop component decomposition algorithms and adopt optimal process planning and decision-making initiatives for multi-process additive manufacturing in order to optimize manufacturing resources, save time, and minimize the geometric complexity. Basinger et al. [60] outlined a feature-based advanced hybrid manufacturing process planning system that uses feature-specific geometric, tolerance, and material data to generate four automated process plans based on user-specified hybrid manufacturing priorities, reducing the manufacturing time. With the application of additive manufacturing in more industrial production applications, the use of HMI technology for the additive manufacturing process chain faces two main challenges.

(1)

There is a lack of a collaborative theory and framework for multiple humans and machines. Although multi-robot cooperation and HMI interactions have been extensively studied, the use of multi-person, multi-machine collaboration in additive manufacturing systems has not. To make collaborative decisions and perform task allocation processes, we must consider factors such as capabilities, resources, and priority levels. We must also study and design efficient information exchange mechanisms to ensure the real-time sharing of critical information between people and machines, such as for task requirement, status feedback, and work progress reports.

(2)

There are still many issues that need to be optimized in terms of software functionality and post-processing for additive manufacturing machines. There is a wealth of knowledge in the field of post-processing that can be applied to additive manufacturing processes, although there is little research on the various post-processing methods that can be used for different additive manufacturing technologies and materials. For example, support needs to be added during the forming process, the software automation and intelligence resources need to be further improved, and the matching of process parameters and materials in the manufacturing process requires intelligent decision-making.

However, industrial production in the context of Industry 5.0 involves personalized characteristics [46], and the purpose of HMI technology is to serve humans and adapt technologies to humans rather than machines. In order to achieve safe human-oriented production, HMI technology alone cannot reflect the role of “humans” in manufacturing systems and cannot clarify how exoskeletons serve humans, thereby achieving safe production outcomes. Industrial tools such as wearable devices with high levels of HMI involvement work in collaboration with the wearer. The complex human movement patterns, intelligent decision-making processes, and variability of the production environment all have an impact on the technical aspects of industrial exoskeletons. Therefore, this study focused on the ever-changing production environment in the context of Industry 5.0.

Table 1. Comparison of traditional technologies with HMI (exoskeleton mechanism design) technology for additive manufacturing processes.

References	Technology Type	Traditional	HMI
[61,62,63]	Mechanism design	Rigid, hard support structure	Flexible skeleton based on ergonomics
		Non-adjustable	Adjustable mechanism adapted to personalized signs
		Traditional mechanical joints	Human biomimetic joints
[64,65,66,67]	Control system design	Static control method without considering environmental feedback	Interactive control method considering wearer feedback
		Multi-parameter complex control method	Simplified control methods for reduced worker stress
		Control methods without insurance measures	Control methods supporting early warnings
[68,69]	Driving design	Electric-driven, constant-stiffness driving method	Intelligent material equivariant-stiffness driving method
		Active driving methods for established programs	Hybrid driving method based on interactive signal perception
[70,71,72,73]	Evaluation design	Evaluation object: operational performance of an industrial body	Evaluation object: the performance of industrial HMI technology in improving human industrial achievements
		Evaluation index system based on physiological indicators and biomechanical evaluation	Evaluation index system considering human–machine trust

shows the differences between traditional technologies and HMI technology.

4.4. Creative Design

Additive manufacturing provides the public with the ability to quickly visualize and manufacture small-batch products, enabling non-professional designers to also be creative. The crowdsourced design approach applies collective intelligence for innovative and creative design and allows collaborative work between people, additive manufacturing information systems, and physical systems through data flow.

The collaborative innovation design model requires the acquisition, analysis, and processing of a large amount of data, which not only covers human ideas and product information such as text, sketches, and computer-aided design models but also includes interaction information between people and products, such as usage scenarios and usage habits. In terms of evaluating the effectiveness of a design, Song et al. [74] proposed that the quantity, novelty, and quality of designs can be used as evaluation indicators for the efficiency of human–machine collaboration efforts in complex drone design tasks. The results indicate that AI assistance is more advantageous for moderately complex targets and has a smaller impact on highly complex targets. However, the use of additive manufacturing for crowdsourcing still presents some challenges, including the lack of a multi-person collaborative design pattern and issues related to the interaction methods and evaluation criteria in collaborative creation efforts between humans and generative artificial intelligence systems.

(1)

Cross disciplinary knowledge integration and the integration of professional and non-professional knowledge are key issues that need to be addressed. The development of a multi-player collaboration model could help users or engineers participate in the creation and adjustment of products in the early stages of the design process. However, how designers from different professional backgrounds and non-professional users can efficiently collaborate remains an open research question.

(2)

Generative design software still needs to be used to address the challenges of generating, modifying, and evaluating printable digital models. While the combination of generative AI and additive manufacturing systems has led to breakthroughs in the analysis and processing of multi-modal data, this approach has also shown great potential for collaborative design. However, it is still extremely challenging to screen and filter valuable solutions from numerous design schemes and improve them. The integration of human experiences and subjective evaluation parameters into evaluation criteria is an urgent problem that needs to be solved.

4.5. Distributed Manufacturing

The combination of digitization and agility in additive manufacturing systems with big data analysis efforts provides a foundation for distributed manufacturing. In the distributed manufacturing model, additive manufacturing companies use additive manufacturing to solve the problems of slow printing speeds and batch production difficulties in order to meet the personalized needs of users in different geographical locations.

The combination of digital additive manufacturing and big data analyses provides opportunities for personalized consumption in both physical and digital spaces through the use of distributed manufacturing. Li et al. [75] achieved self-organizing team collaboration in HMI systems by introducing the cognitive manufacturing paradigm of informatics. This can be used to evaluate various HMI models in decentralized factories and aggregate them into a universal knowledge representation model. Based on existing experiences, this can then be used to allocate the best self-organizing tasks for the HMI system, for improved collaboration efficiency and flexibility, and to achieve large-scale personalization. In terms of the military applications, battlefield and other combat environments rely on rapid solutions. The main challenges when using this model are to obtain dynamic data based on application scenarios, understand the requirements, and complete the online planning of manufacturing point orders and logistics requirements. Multi-objective optimization and adaptive planning algorithms for complex systems based on industrial Internet- and AI-aided decision-making provide opportunities to solve these problems. In this model, the roles of participants in the production process, such as workers and designers, are also changing. The main function of workers is to lead the production strategies and manage the implementation of self-organized production processes. Due to the widespread availability of network and mobile real-time information, traditional fixed workplaces have become less important. Regarding complex problems, workers will assume the role of creative problem solvers [76]. For distributed manufacturing, the key challenges in the future include two aspects—technology and value.

(1)

From a technical perspective, improving the adaptability and resilience of the manufacturing process is a key challenge. In the manufacturing process, the dynamic changes in the supply chain and transportation bring about uncertainty regarding human–machine cooperation. The question of how best to coordinate human–machine cooperation efforts in different time and space scenarios to ensure the stability and reliability of the manufacturing process needs to be reconsidered from the perspective of system design.

(2)

From a value perspective, redefining and creating value from human labor is a key challenge. To achieve a people-oriented value system, the evolution of worker roles and changes in the workplace require a redefinition regarding the labor value and mode of labor of individuals in the production process.

4.6. Energy Conservation and Emissions

HMI additive manufacturing is playing an important role in achieving the goal of energy conservation and emission reductions, as well as in promoting the development of a circular economy. On the one hand, additive manufacturing can be used to reduce raw material waste outputs and achieve energy-saving and emission reduction goals during product manufacturing [77], repair, and remanufacturing processes through the use of layer-by-layer manufacturing and lightweight design methods. For example, in the aerospace field, the application of additive manufacturing technology to turbine blades has led to improved efficiency levels and reduced costs. Meanwhile, the reusability, health, and carbon emission management of additive manufacturing materials are new opportunities for the future [78]. For example, the recycling and reuse of polymers in additive manufacturing have led to improved productivity and sustainability in the circular economy [79]. On the other hand, AI technology can be used to assist designers in making decisions on the design, manufacturing, and assembly of complex products together with smart machines integrated into the IoT, providing enormous potential for energy conservation and emission reductions. On the design side, knowledge-based models can be used to assist designers in designing solutions that meet manufacturability requirements [80], thereby reducing scrap rates. On the manufacturing side, knowledge engineering can be used to help designers automatically generate process plans [81] and machine learning algorithms can be used to optimize process parameters to achieve the parallel optimization of production, energy consumption, and product quality efforts [82]. On the production side, machine learning can be used to assist designers in pre-manufacturing efforts and in planning [83], evaluating, and controlling the product quality [84]. Therefore, human–machine collaborative additive manufacturing provides a new way to achieve energy conservation and emission reduction goals and promote the circular economy in the development and production of complex products.

5. Conclusions

This paper has proposed the concept and framework of HMI additive manufacturing for Industry 5.0 based on the theory of HCPSs. Through a people-oriented perspective, this study investigated the human–machine collaboration relationship in HMI additive manufacturing for Industry 5.0, pointing out that the use of HMI additive manufacturing for Industry 5.0 is centered on human beings in the context of additive manufacturing. Through the use of HMI technology, the advantages of humans and machines are fully utilized to achieve the synchronous improvement of human and machine capabilities, better meet user needs, and promote sustainable social development. This article has proposed a product development framework for HMI additive manufacturing, which focuses on the characteristics of diversified, parallel, and dynamic human–machine collaborative relationships in the development of additive manufacturing. Additionally, taking typical applications at the product layer, economic layer, and ecological layer as examples, this article has elaborated the different roles and effects of human and additive manufacturing systems through different interaction methods in different application scenarios, introducing specific technical implementation paths and research results. HMI additive manufacturing for Industry 5.0 can be used to achieve more friendly and convenient human–machine collaboration relationships at various stages of product development, improve the capabilities of humans and machines, and meet personalized human needs. At the same time, the development process of HMI additive manufacturing should also treat sustainability and resilience as important considerations. This article has explored the conceptual framework and typical application scenarios of HMI additive manufacturing for Industry 5.0, although there are still some problems and challenges relating to the application of this framework.

(1)

Deep collaboration between humans and machines: On the basis of existing human–machine collaboration relationships, deep HMI technology aims to maximize the advantages of both humans and machines in diverse human–machine collaboration modes. This poses new challenges regarding the intelligent perception capabilities of machines, secure interactions between humans and machines, and data processing. This not only requires the AI technology to have strong computing power levels to complete perception and reasoning tasks but also to have the ability to understand humans in order to achieve deep and unobstructed collaboration with them in more diverse scenarios.

(2)

Data-driven traps: While AI-based data-driven methodologies enable designers to fully explore the value of many forms of data in product creation, they can also aid in decision-making, helping to accomplish tailored product design and manufacturing goals. Data arrogance, algorithm evolution, and potential motivation issues will present additional hurdles to HMI additive manufacturing as the parallel product creation processes become more complicated.

(3)

Sustainable manufacturing and manufacturing resilience: The use of large-scale additive manufacturing leads to increased production uncertainty due to dynamic human, machine, and environmental factors. Additionally, the use of additive manufacturing through HMI requires sustainable manufacturing and resilience efforts. The remanufacturing workflow must account for intelligent design, manufacturing, inspection, and assembly initiatives, as well as complicated product production and supply chain flexibility and agility efforts.

The concept of HMI technology in additive manufacturing is highly compatible with Industry 5.0, and the research objective is to continuously move towards a human-oriented manufacturing process. By finding the convergence point of efficient and collaborative work between machines and humans, we can fully leverage the capabilities and advantages of machines and humans to meet their personalized needs. The collaboration between humans and additive manufacturing machines relies not only on the perception, computation, and cognition of machine intelligence systems but also on human intuition and creativity. The advancement of technology has not separated people from the problems they face in society but rather enables people to better understand the world, recognize problems, and have the opportunity to solve these problems. In the future, user research, usability testing, and iterative design efforts will be essential in creating seamless additive manufacturing and system operator experiences. AI-powered analytics can be used to uncover patterns, optimize resource allocation efforts, and boost efficiency by analyzing massive production data. Robots can help with material handling, part inspection, and repetitive duties, allowing humans to work on creative problem-solving jobs. The use of designs for recyclability, eco-friendly materials, and closed-loop material recovery and reuse systems could lead to improved product sustainability throughout the lifecycle.